# figure size in px
facet_col_wrap = 6
if facet_col_wrap==6:
width = 1100*2
height = 300*2
elif facet_col_wrap==3:
width = 500*2
height = 600*2
else:
assert False
#col_options = {col_name:pd.unique(df_long[col_name]).tolist() for col_name in consistent_cols}
#display(col_options)
def plot_combined_images_plotly(order_current, df_long=df_long, do_save=True, save_scale=None, save_extensions=None, n=None, annotate_extra_info=True):
if order_current is None:
order_current=0
#order_current=order_current+1
df_sampled = df_long.loc[df_long.Order==order_current]
while df_sampled.shape[0]==0: # skip ahead for empty order_currents
return False
#order_current=order_current+1 if order_current < df_long.Order.max() else 0
#df_sampled = df_long.loc[df_long.Order==order_current]
data_row = df_sampled.iloc[0] # get first row
base_row = df_sampled.loc[df_sampled.mmHg==0].squeeze()
base_folder = os.path.join("data/01_raw/photos/", ("glove" if data_row["Material Type"]=="Glove" else "condom/orig"))
images = np.array([skimage.io.imread( os.path.join(base_folder, filename)) for filename in df_sampled.Filename])
# rotate
# images = np.rot90(images, k=1, axes=(1,2))
dim = images.shape[1:] # should be 3 values, with the last being color dim
dim_small = tuple([int(dim[0]/n), int(dim[1]/n), *dim[2:]])
if data_row['Material Type'] == "Glove":
data_row_str = [
f"'{data_row['Method']} finger' method" ,
f"with {data_row['Size']}. {data_row['Material'].lower()} glove," " " f"Trial #{data_row['Trial']}"
]
data_row_elem_str = [
f"'{data_row['Method']} finger' method" " " f"at {data_row['mmHg']}mmHg",
f"with {data_row['Size']}. {data_row['Material'].lower()} glove," " " f"Trial #{data_row['Trial']}"
]
elif data_row['Material Type'] == "Condom":
data_row_str = [f"{data_row['Material']} brand condom," " " f"Trial #{data_row['Trial']}"
]
data_row_elem_str = [
f"{data_row['Material']} brand condom" " " f"at {data_row['mmHg']}mmHg," " " f"Trial #{data_row['Trial']}"
]
elif data_row['Material Type'] == "None":
data_row_str = [f"No material," " " f"Trial #{data_row['Trial']}"
]
data_row_elem_str = [
f"No material" " " f"at {data_row['mmHg']}mmHg," " " f"Trial #{data_row['Trial']}"
]
def ind_to_row(index):
return 6//facet_col_wrap - ((index // facet_col_wrap)) # // is integer division (divison without remainder)
def ind_to_col(index):
return (index % facet_col_wrap) + 1 # % is modulus operator
fig = px.imshow( images[:, ::n, ::n, :], binary_string=True, origin="lower", aspect="equal",
facet_col=0, facet_col_spacing=0.0,
facet_col_wrap=facet_col_wrap,
#category_orders=category_orders,
labels={**labels,"facet_col":"mmHg"},
#title = " ".join(data_row_str)
)
# Add annotations
#for index, data_row in df_sampled.iterrows():
for index, (index_of_all, data_row) in enumerate(df_sampled.iterrows()):
x0=(dim[1]-(data_row.wd_start))/n;
x1=(dim[1]-(data_row.wd_start+data_row.wd))/n
y0=(data_row.ht_start)/n;
y1=((data_row.ht_start+data_row.ht))/n
# Draw box around labeled area
fig.add_shape(
type="rect",
x0=x0, x1=x1,
y0=y0, y1=y1,
xref='x', yref='y',
line=dict(
color="red",
width=4,
dash="dot",
), opacity=0.5,
row=ind_to_row(index), col=ind_to_col(index),
)
# Draw vertical lines around the labeled width (verlaps with box above)
for ind, x in enumerate([x0, x1]):
fig.add_vline(
x=x,
line=dict(color="red",width=4,dash="dot"),
opacity=0.75,
row=ind_to_row(index), col=ind_to_col(index),
annotation_position=["top right","top left"][ind],
annotation=dict(text=[
f"<b>{data_row.wd/base_row.wd:.1%}</b><br>of baseline ",
f" <b>{data_row.wd:.0f}px</b><br>view width"][ind],
font_size=14,
bgcolor="rgba(255,255,255,0.4)",
),
),
fig.layout.annotations[index]["text"] = f"<b>{data_row.mmHg}mmHg</b>"
#fig.layout.annotations[index]["text"] = f"View Width: {data_row.wd:.0f}px ({data_row.wd/base_row.wd:.1%})"
#fig.update_xaxes(title=f"At {data_row.mmHg}mmHg", row=ind_to_row(index), col=ind_to_col(index))
# Add filename and other info annotation
if annotate_extra_info:
for index, (index_of_all, data_row) in enumerate(df_sampled.iterrows()):
fig.add_annotation(
x=1, y=1,
xref="paper", yref="paper",
text="<br>".join(data_row_elem_str) + "<br>Filename: " + data_row.Filename + "<br>Size: " + str(dim[0]) + "x" + str(dim[1]) + "px" + ("" if n==1 else f" (reduced {n}x to display here)"),
xanchor="left", yanchor="bottom", align="left", # align only matters if multiline. 'anchor' arguments actually change the position within the graph
font=dict(size=10, color="blue",family="Courier"),
row=ind_to_row(index), col=ind_to_col(index),
showarrow=False,
)
#fig.update_traces(hovertemplate="x=%{x*" + str(n) + "} <br> y=%{y*" + str(n) + "} <br> color: %{color}")
ycoords, xcoords = np.meshgrid(np.arange(dim_small[0])*n, np.arange(dim_small[1])*n, indexing="ij")
customdata = np.dstack((xcoords, ycoords)) # shape=(height, width, number of data values)
#fig.update(data=[{'customdata': np.dstack(np.meshgrid(np.arange(dim_small[0])*n, np.arange(dim_small[1])*n, indexing="ij")),
# 'hovertemplate': "x: %{x} <br> y: %{y} <br> xn: %{customdata[0]} <br> yn: %{customdata[1]:.3f} <br> z: %{z} <br> color: %{color}<extra></extra>"}])
fig.update_traces(
customdata=customdata,
hovertemplate="x=%{customdata[0]} <br>y=%{customdata[1]} <br>color=%{color}"
)
fig.update_layout(
font=dict(
family="Arial",
size=20,
color="black",
),
title={
"y":1,
"x":0.5,
"xanchor": "center",
"yanchor": "top",
"font":dict(size=20)
},
paper_bgcolor="#F9F9F9",
margin=dict(l=0, r=0, t=40, b=0),
width=width, height=height,
#dragmode="drawopenpath",
newshape_line_color="cyan",
)
fig.update_xaxes(showticklabels=False) # , title_text=[f"At {mmHg}mmHg" for mmHg in df_sampled.mmHg]
fig.update_yaxes(showticklabels=False)
fig.for_each_trace(lambda t: t.update(name = f"At {int(t.name)*40}mmHg",
#legendgroup = f"At {t.name*40}mmHg",
#hovertemplate = t.hovertemplate.replace(t.name, f"At {int(t.name)*40}mmHg")
)
)
fig.show()
if do_save:
save_plotly_figure(fig, file_name=f"Trial {order_current}) {' '.join(data_row_str)}, n={n}", save_in_subfolder="Each Trial - Plotly",
scale=save_scale, extensions=save_extensions )
